A distributed control system in which plural control devices are mutually connected to one another through signal lines have been hitherto used in various fields. For example, a vehicle is equipped with a distributed control system for controlling control targets such as an engine, a transmission, etc. by an engine control device, a transmission control device, etc. while data are delivered among these control devices.
The distributed control system as described above broadly uses a method of serially transmitting plural kinds of data with a pulse train signal for the communications among the control devices. As the communication method as described above are known a method of alternately transmitting data and header data representing the attribute of the data concerned (see JP-B-7-71068; patent document 1), and a method of allocating different kinds of data to the on-timing interval and the off-timing interval of two pulses and then transmitting the data (see Japanese Patent No. 3389643; patent document 2).
However, when one kind of data D are transmitted according to the method disclosed in the patent document 1, as shown in FIG. 20A, the time corresponding to one period is consumed for each of transmission of the data D and transmission of the header data H, that is, totally the time corresponding to two periods (2t) is consumed. Therefore, there is some limitation to enhance the communication efficiency, and thus this method is undesirable.
Furthermore, according to the method disclosed in the patent document 2, as shown in FIG. 20B, the time consumed for the transmission of each of the two kinds of data D1, D2 is suppressed to one period, and thus the communication efficiency can be more enhanced by the method disclosed in the patent document 2 than that by the method disclosed in the patent document 1. However, as a result of diligent studies of the inventor of this application, it has been found that the following problem would occur if the data representing the state quantity of the control target and the data representing abnormality of the control target are transmitted as the data D1, D2 respectively to secure safety of the distributed control system. The problem resides in that the data D2 representing abnormality of the control target is needed only when the abnormality concerned occurs, and thus the data D2 becomes substantially unnecessary dummy data when the control target is normal. That is, when the control target is normal, one kind of data D1 is transmitted equivalently by consuming the time corresponding to the two periods (2t), and thus the enhancing effect of the communication efficiency is hardly achieved.